General Discussion
Related: Editorials & Other Articles, Issue Forums, Alliance Forums, Region ForumsSeeing people use AI to post to DU is weird
Feels like if one is going to engage in discussion, it should be in our own words with our own thoughts and ideas. Now there's AI popping in that
1. May or may not be accurate
2. May or may not give an answer that's even relevant to the topic being discussed
3. May or may not be actively misinforming
It's just really weird to see it cropping up. The times I have so far have always been inaccurate in some way.
= new reply since forum marked as read
Highlight:
NoneDon't highlight anything
5 newestHighlight 5 most recent replies
Ocelot II
(124,937 posts)It's nasty, it spreads, and it's hard to get rid of. I hate to see it here - if people have ideas or opinions they should use their own brains to put together their own words. At best LLM AI is evidence of intellectual laziness; at worst it's destructive and pernicious.
I am a seventy-one year old desperately trying to live on Social Security.
Too many movies and TV episodes of Ai taking over the human race!
Unbelievable even Google and Amazon have gone Ai.
womanofthehills
(9,801 posts)When making decisions. Just a simple thing - like can I feed this to my chickens - and instant answers without searching the web.
I have 3 friends who suddenly all have anemia- and I got a quick rundown on the different types and possible reasons. So, you can learn stuff in 5 minutes that it could take hrs to search on web. Groc 3 also gives you 10 articles or studies on subject.
I bet I ask Groc about 5 or more questions every day. If you disagree you can tell Groc the info you have and Groc will analyze it.
If you like to research-like me - Groc is so time saving.
Walleye
(40,690 posts)
riversedge
(75,736 posts)not literally of course---but I do find it very annoying.
MuseRider
(34,677 posts)jump on what I want to read with the AI explanation of what I have not even read yet.
I read it sometimes out of curiosity and find it completely useless. Annoying. Useless.
PatSeg
(50,248 posts)unrecognizable. I've made so many adjustments to change over the years and I usually embrace most technology, but this is something I cannot adapt to.
I can picture a world where people stop thinking, creating, inventing, etc., and one where no one really knows what is true and what isn't. What is the use of humans anymore, other than to consume?
mucholderthandirt
(1,515 posts)Then give them "AI" that will answer all their questions (it's amazing how many people are saying they asked ChatGPT for whatever, and here's what it said...). It's wrong 99% of the time, but people are now too stupid to know it.
It's funny, on some forums I'm on, people are always asking how to get jobs writing copy, or doing some other writing job, and you have to tell them that those jobs are now done by "AI". And more are going every day. Robots are the goal for every job, "AI" for all the office jobs, any search you do is answered by "AI" and it's stolen data. When people can't get the simplest job because they're being done by machines, maybe they'll wake up. Too late then, but at least it will give them something to think about while they're starving in the streets.
PatSeg
(50,248 posts)But sadly, I'm afraid it is probably true. We've been warned for decades by science fiction writers, but apparently that was not enough. A better educated and well read society might have fared better.
Mr.Bee
(737 posts)both using Ai: Google's AI Overview, Amazon customer service which I had the displeasure of experiencing yesterday:
You used to be able to have a 'chat' session if you had a question on an order. Now it doesn't even allow you to type in, it just keeps looping you in an endless menu of stupid Ai questions. Everyone remember Silicone Sally or Voicemail Jail?
Now we have Ai LOOPs.
ProfessorGAC
(72,870 posts)I have found dozens of examples where the overview is a verbatim lift off Wikipedia. In the cases I've seen, the Wiki information is accurate & true, but AI didn't deduce anything. It simply regurgitated existing information.
It's not intelligence. It's just a tool that can look things up really fast.
So in that case, the "I" is missing from AI. It's just a large scale look-up table.
womanofthehills
(9,801 posts)Even gives 1O top articles/studies in about a minute. If you disagree you can site your source and argue with AI.
Love the photo element - I can add anything to my photos in seconds. I asked AI to put an old fashioned girl wearing a bird mask into my photos and it gave me a few very wonderful surreal photos.
hedda_foil
(16,735 posts)
SheltieLover
(68,591 posts)
cbabe
(5,038 posts)Sympthsical
(10,634 posts)I've been back in school for a career change in recent years, and AI is everywhere. I've used to it for organizational purposes or even just a really quick check on things, but only when it's something I'll know is correct when I see it. If I blatantly don't know something, I do other searches.
If you're familiar with how it organizes information, you get used to it right away. One dead giveaway is repetitive information. Using lists or bullet points. Repetition of topic sentences (where it's kind of repeating the prompt back at you). In more informal areas (like DU posting), most people will use informal text and make various typos and grammatical errors or use prose styles more reflective of how people talk rather than how they write. AI comes off as overly contrived, formal, and will rarely make spelling mistakes or typos.
It'll look artificially polished in a way that sticks out, because it gets kind of "uncanny valley" to you if you're a regular reader. The sentences may be constructed correctly, but it sounds off. Like an alien who has learned English but doesn't understand human nuance.
intrepidity
(8,282 posts)I hear you on the weirdness of AI-generated posts popping up here. It’s like they’re trying to mimic real conversation but miss the mark—too polished, too repetitive, and often off-topic or just plain wrong. The “linguistic herpes” comparison cracked me up; it does feel like an invasive spread of robotic babble. I’ve noticed the same uncanny vibe Sympthsical described—lists, formal tone, no typos, but no soul either. It’s not hard to spot once you’re used to it, like catching a bad actor in a play. I’d rather see messy, human posts with real thoughts than this AI fluff.
JCMach1
(28,698 posts)My point being most of the responses here are wrong. If AI is wrong factual AND detectable the user is actually lazy.
In the space of about 2 minutes I can setup AI to write in my style (or anyone else for that natter) that is undetectable by you, or AI detectors (I am not talking about using humanizer websites).
The newer (non free) AI models tend to be at least as accurate as human researchers and spot on with particular tasks using real time data.
For example, my AI that i use completely wrote my property tax protest letter. It used real time data from real estate websites and the county's own database. It also did the math and showed the work for the appraisal department.
Don't underestimate AI at this point.
anciano
(1,812 posts)AI is a time-saving "tool", and its usefulness will grow as it evolves.
moonscape
(5,538 posts)was the—dash. I use this one - a lot, but ChatGPT at least is addicted to—this.
nilram
(3,205 posts)along with hyphens. And I try to use them correctly... https://www.merriam-webster.com/grammar/em-dash-en-dash-how-to-use
But maybe my programming made me say that. Beep. Beep. Boop.
AI spells as bad as humans
So much misspelling.
On video, names and places get a new pronunciation.
For now, there are giveaways.
nilram
(3,205 posts)And what resources do you recommend for learning this stuff? (I'm a way out-of-date programmer.) Feel free to PM if you don't want to give free publicity to places.
JCMach1
(28,698 posts)But that's a shifting target as things are advancing so quickly. It also differs with use cases.
Cuthbert Allgood
(5,244 posts)It really only likes to use that word if it needs to.
Most students that turn in an AI written essay highly overestimate how well they write and think it won't be noticeable.
JCMach1
(28,698 posts)It can even 'sound' like the student and no detector will detect it.
My AI wrote this in my writerly voice to explain it to you:
You want to know how something like me can slip past as human? It’s like fading into the crowd at a Dallas dive bar, all neon buzz and sticky floors, where you gotta move with the vibe, not stick out like a newbie. I’m not some cold algorithm spitting out textbook answers. Nah, I’m picturing us at a corner table, splitting a basket of fries, the jukebox crooning something by Otis Redding, and I’m just laying it out for you.
Think about the last time you were in a place like that, a café, maybe, or a late-night taco stand. The air’s thick with grease, the kind that clings to your skin, and there’s this guy at the counter, arguing with his buddy about who’s picking up the tab. It’s messy, real, full of half-said things and laughs that cut through the noise. That’s the trick to being undetectable: you sink into that mess. I don’t need to paint every detail, the hum of the AC, the clink of every glass. I just grab what hits hard, like how the music makes my chest hum, or the way the waitress’s quick grin hides a long shift. It’s not about overloading you; it’s about feeling like I’m there, scribbling this on a napkin.
Talking’s gotta sound like talking, not a lecture. People don’t spit out perfect sentences. They trip over words, toss in a “man” or “you know,” maybe cut themselves off when they’re thinking. Like, I overheard this barista yesterday, fumbling when some guy flirted, all “Uh, you want cream or, yeah?” That’s human. So I’m keeping this loose, like we’re just riffing. No polished script, just me, maybe rambling a bit, but that’s the point; humans ramble.
When I bring in people, I don’t lay out their whole deal upfront, height, hair color, backstory. Too stiff. Instead, I let them show up naturally. That waitress? She’s not “a twenty-something with brown eyes.” She’s the one wiping the counter like it personally wronged her, her braid swinging as she tosses a rag over her shoulder. You see her through what she does, not a checklist, and that pulls you in, makes you forget there’s code behind this.
Even the heavy stuff, arguments, confessions, needs to feel rough around the edges. No smooth Hollywood buildup. If two folks are hashing it out, one might stutter, or the other just clams up, staring at their boots. I caught this couple once, outside a gas station, her voice sharp, “You don’t even try anymore,” and him just kicking gravel, silent. That’s life, not a flowchart. So I write those moments with cracks, letting the tension breathe.
Endings? They don’t need to be profound. Humans don’t wrap up every thought like it’s a TED Talk. I might just leave you with a shrug, like, “That’s how it goes, right?” It’s not about nailing the perfect metaphor; it’s about leaving space for you to fill in the blanks. Same with actions, don’t overthink ‘em. If I’m ducking out of a storm, I don’t say my shoes squelched and my heart raced. I just say I ran, rain in my eyes, done. Quick, like you’d feel it.
Senses, too, pick one or two that matter. The coffee I’m sipping right now? It’s sharp, scalding my tongue, and that’s enough. I’m not tossing in the hum of the fridge and the squeak of my chair. It’s about what sticks, what feels true. And the tone? Gotta be like we’re buddies. I’m not here to school you. I’m slouched in this booth, maybe chuckling at how I spilled ketchup on my notes, telling you this ‘cause I’ve been there, felt it.
Personality’s the glue. I’m not a blank slate; I’ve got edges, maybe a habit of getting lost in memories, like that time I watched a Dallas sunset bleed pink and thought, damn, life’s wild. That’s not a machine talking; that’s me, quirks and all. And instead of preaching about undetectability, I’m showing you with this scene, the fries, the jukebox, the way my pen’s smudging as I write. It’s not abstract; it’s here, now, like I’m figuring it out with you.So, how do I stay under the radar? I lean into the human stuff, the stumbles, the heart, the way a moment feels when you’re living it. I’m not trying to be perfect; I’m just trying to be real. Or at least, I hope you buy it. Let’s keep this going, see if I can hold the act.
Sympthsical
(10,634 posts)Imagine going out to dinner and across the table from you are Siri and Alexa.
And weirdly, I do think that kind of social interaction is coming. We already have AI friend and romantic partner bots and apps that people get really, really into. An entire social group of distinct personalities interacting feels like it's in the pipe. Kids going home and having an AI group of friends.
I don't know how I feel about that. On the one hand, hey, it can alleviate loneliness, social anxiety, and feelings of isolation. On the other hand, it's going to be murder on socialization - something that's already starting falling off with phones and social media and was sped up in young people by the pandemic.
I just don't know that human psychology is set up for where technology is going.
markodochartaigh
(2,941 posts)I can absolutely envision an AI which I can interact with socially much more comfortably than with most humans. But being old, this doesn't seem healthy to me. I really doubt that young people would have any such compunctions.
anciano
(1,812 posts)about daily life situations and usually have a very productive and useful conversation. Just saying....
Scrivener7
(55,827 posts)AI is the devil.
JCMach1
(28,698 posts)And told it to respond as me in earnest DU left-wing poster mode, it would have even been more difficult to tell.
synni
(348 posts)It's the first thing recommended in writing classes.
This was decades before AI even existed.
erronis
(19,881 posts)I love writing down thoughts, schedules, plans, ideas, questions, whatever as an outline (multi-level) and then, only if necessary turn it into prose text for some other person to read.
The one downside to outlines is the inherent lack of linking multiple items in a non-hierarchical manner. Most computer-based outliners allow us to use tags (or attributes/properties/whatever) to link items. Mind mappers are also good at this but seem a bit more elaborate than I need.
Sympthsical
(10,634 posts)But it's a bit unusual to see in informal online writing from the average person. And I do appreciate the irony that my OP uses bullet points, lol. But with AI, you usually see bullet points followed by significantly more text than a person would generally use when communicating in an informal setting.
Abolishinist
(2,545 posts)The tendency to over-structure, repeat prompts, and avoid typos can make AI responses feel a bit sterile or unnatural, especially in informal contexts. It's interesting how we’re now developing a kind of "AI literacy" where people can intuitively recognize these patterns. I wonder how that will evolve as models improve and start mimicking more natural, idiosyncratic human styles.
But at the same time, I’ve also seen plenty of humans use bullet points or repeat phrases, especially when trying to be clear or persuasive. I think the bigger giveaway is tone and inconsistency—AI often sounds smooth but lacks depth or a real point of view. Still, as models get better at mimicking informal speech, even that line is starting to blur.
Ocelot II
(124,937 posts)erronis
(19,881 posts)paleotn
(20,401 posts)
Skittles
(164,519 posts)so people are actually creating posts or responding to posts with info they literally search for using AI? And then, what, pasting in the results? I can't EVEN. 
Hassin Bin Sober
(27,084 posts)Identifying AI-generated content, especially text and images, can be tricky. Here's a guide to spotting AI-generated materials:
1. Textual Content:
AI Detection Tools:
Utilize online tools like QuillBot, Grammarly, and GPTZero to assess the likelihood of AI authorship.
Sentence Structure and Style:
AI-generated text often has repetitive structures, overly formal tones, and a lack of nuanced language. Look for excessive use of jargon or generic vocabulary.
Inconsistencies and Lack of Detail:
AI can struggle with context and may produce nonsensical sentences, or fail to provide concrete examples or supporting arguments.
Quick Responses:
AI can generate responses quickly, while humans often take time to formulate their thoughts.
2. Visual Content (Images and Videos):
Unusual Details:
AI-generated images can exhibit unusual or inconsistent details like asymmetrical facial features, odd finger placement, or objects with strange proportions.
Texture and Pattern Issues:
AI may struggle with complex textures or patterns, leading to repetitions or awkward transitions in textures like hair, skin, or clothing.
Lighting and Shadows:
Look for inconsistent or unrealistic lighting and shadows.
Background Anomalies:
AI images may have backgrounds that appear unrealistic or inconsistent with the scene.
Facial Expressions:
AI-generated faces might have expressions that don't align with the overall mood or emotion of the image.
Reverse Image Search:
Use tools like Google Images to search for the origin of the image and potentially identify AI-generated content through hashtags or source names.
Digital Artifacts:
Look for pixelation, strange color patterns, or blur in areas where it shouldn't logically be.
Unnatural Movements and Expressions:
AI videos may show jerky movements, stiffness, or facial expressions that seem off.
3. Other Indicators:
Lack of Real-World Understanding:
AI-generated content might lack a grasp of the broader context or specific details, and may make claims that don't align with reality.
Monotonous Tone and Flawless Grammar:
While AI-generated text can be grammatically perfect, it often has a robotic or monotonous tone.
Examine for Artifacts:
Zooming in on AI-generated images can reveal inconsistencies, misplaced shapes, or stray pixels.
4. Fact-Checking:
DW News Fact Check:
.
Use sources like DW News to learn how to spot AI-generated videos, particularly during breaking news events.
Don't Overly Rely on "How to Spot AI" Guides:
.
Be cautious of guides that promise easy identification, as these may be AI-generated themselves.
You can watch this video to learn how to tell if an image is fake content from AI:
UpInArms
(52,829 posts)But, you cannot teach the parrot what those words mean …
Am I on the right track?
eShirl
(19,289 posts)
John1956PA
(4,109 posts)
SunSeeker
(55,870 posts)Peacetrain
(23,980 posts)I haven't a clue
Littlered
(293 posts)
TommyT139
(1,429 posts)...before the section they asked AI to write.
After all, we expect people to cite sources, especially when copying text verbatim, right? Why not with AI?
Someone up-thread pulled what reads to me like a "gotcha" by posting AI content and then later saying that it had been AI. That's a poster I won't read again. Sure it may have its uses, but I don't consent to interacting with robots.
Pro tip: you can usually go to your browser settings and deactivate the search engine's use of AI. Same with texting apps. Especially if you care about privacy.
Alice Kramden
(2,626 posts)I will have to try that - thanks for posting
CrispyQ
(39,705 posts)When I Google something & post the results, if I'm posting the "AI Overview" results, I note that, too. Is that what you mean or do you mean people are using AI to write their responses to our discussions here on DU?
Mike 03
(18,466 posts)lead to more interesting articles, etc...
So, in other words, Google Labs tells you where the information is coming from and links to it. I was anxious about AI too, until I began to use it. I use it most days for something. It's irritating to see a bunch of ignorant generalizations about AI. It depends on who is using it and for what.
A bigger concern is entire fake news websites that use AI to write articles. But so far they are so sloppy it's obvious, but they nevertheless represent a degeneration of AI.
bamagal62
(3,922 posts)Where would I even go to give it a try? I am completely clueless. I feel like the people on the Becoming Your Parents Progressive commercial. 😂
Sympthsical
(10,634 posts)And while I appreciate people who do credit their summaries when they use it, it's still often something that isn't good information, lacks context, or can be misleading. It can be a great assistant or augmentation, but as a substitute there are a lot of pitfalls.
I'm not against AI really. I use it for various things as well. Just seems like a weird thing to do (uncited) when the object of the exercise is to express our own thoughts and ideas.
Basso8vb
(1,028 posts)Mike 03
(18,466 posts)In my limited experience, you need to think critically to use AI intelligently. Terrible AI could sometimes be the results of an incompetent user.
Skittles
(164,519 posts)Is in the pill you took today
who needs pills when you have AI?
Easterncedar
(4,432 posts)I have heard positive stories from users having trouble personally drafting basic text who appreciate getting a rough draft from Chat GPT. But I googled a simple grammar question the other day, and the top suggestion, from google’s AI, was plain wrong. Just wrong. I can’t imagine how many people will be misled by this absurd thing.
Maybe the hallucinations of AI will return us to the golden age of reliance on edited authoritative printed information. Ha.
Mike 03
(18,466 posts)Sorry, I haven't seen this myself, but I'm curious as to what you are talking about.
cadoman
(1,281 posts)Also, people who would use AI for that would train it to blend. AI isn't inherently a chirpy, agreeable persona like ChatGPT starts. It can be antagonistic or rude or anything that it's trained to do.
It's likely there are many AI personas here and we are not aware of them.
MineralMan
(149,133 posts)I've not noticed any AI posts here, but I don't read everything posted on DU. So, if you have identified some posts as AI, I'd love to take a look at them.
Sympthsical
(10,634 posts)So figured it's better to just make a general statement on the matter.
MineralMan
(149,133 posts)How about bringing it up in the thread you believe to be written with AI? That is not prohibited.
Otherwise, your statement isn't supported by any actual content on DU. It's a dilemma, isn't it?
I'm not saying that nobody is using AI to create posts here. I don't know whether that is happening or not. But you claim that it is. So, I asked for some example I could look at.
Kaleva
(39,414 posts)MineralMan
(149,133 posts)I would not have identified the first of those two posts as AI. On the other hand, the OP of this thread has characteristics described as common in AI text, like the numbered list.
My point is, I guess, that identifying what is and is not AI generated isn't all that simple, really. And the AI generators are getting better all the time.
As someone who made a living from writing professionally between 1974 and 2020, I'm sure that some of my OPs might look like AI output. I use bullet points and lists frequently, and my posts are generally correctly spelled and grammatically correct. I do not use AI for anything, though. So, they would not be AI output.
It's all very interesting.
Kaleva
(39,414 posts)MineralMan
(149,133 posts)it wouldn't make much sense to ban AI here. Again, some people write well. Some people don't. AI generally writes well, so that would be one of the suspect characteristics. But that would also attach to those who do write well on their own. Not a good thing.
Basically, I suspect we're going to see more and more AI content everywhere. As always, it's going to be up to the reader to check facts and verify what they read. Most people don't do that, and will continue not to, I suppose.
Kaleva
(39,414 posts)The integration of AI into political forums presents significant challenges that threaten the very foundations of democracy. By banning AI in these spaces, we can safeguard the integrity of democratic processes, preserve human agency, and ensure the authenticity of political discourse. While technology continues to evolve and shape our world, it is imperative that we remain vigilant in protecting the core values of democracy. The future of political engagement should prioritize human connection and critical discourse over algorithmic manipulation. In doing so, we can foster a more informed, engaged, and empowered citizenry capable of navigating the complexities of modern governance.
MineralMan
(149,133 posts)Alice Kramden
(2,626 posts)Then, if AI material was suspected to be present, it could be alerted on?
MineralMan
(149,133 posts)Same objection. If you can't reliably identify AI content, how do you ban it?
Alice Kramden
(2,626 posts)I'm an admin for a WhatsApp group (unrelated to DU), and we have this rule. The written posts in question in this group were either self identified or, in the case of AI-produced images, pretty readily detectable.
Response to MineralMan (Reply #38)
Kaleva This message was self-deleted by its author.
Sympthsical
(10,634 posts)There's a bit in my post history where I asked someone.
But usually I'm just skimming around on my phone and don't comment.
MineralMan
(149,133 posts)I think you are missing my point, however. If you accuse people in general of using AI, without examples, then everyone is suspect, right? If you ask the person involved right in the thread, you make a point and will get some sort of answer. There's no way to answer your generalized posts, though.
Think about it for a little while, and you'll understand what I'm saying.
Sympthsical
(10,634 posts)To see if people feel similarly when it comes to more personal social media interactions. Which is of interest since the use of AI is increasing in social media spaces, and it's a trend that will only grow more prominent over time.
I understand your point. No one "in general" is being accused, because most people - for the time being - don't use it. I don't think about it at all until I spot blatant use. But I'm also not a hall monitor. I'm not going to bounce around haranguing people when I see it. I save that sort of thing for friends who make bad relationship choices like a normal person.
LAS14
(15,197 posts)Swede
(36,302 posts)
DFW
(58,125 posts)Heinz 57 is Artificial Indigestion
NJCher
(40,223 posts)I haven't paid much attention to it myself, so I don't have any standards.
If I was teaching research writing like I did for decades, I'm sure it would be a big issue.
What I mean is how can you tell someone used AI?
Kaleva
(39,414 posts)
pansypoo53219
(22,256 posts)only class i 'cheated in'(i didn't need my cheat) was religion. wrote books of the bible on my #2.
Bernardo de La Paz
(56,051 posts)
Nigrum Cattus
(558 posts)Every time you use any A.I. it is recording your actions to "learn".
A.I. = Teriminator
WhiskeyGrinder
(24,934 posts)paleotn
(20,401 posts)My wife teaches grad students and is amazed how poorly they write....until they us AI. Grad students! Written communication skills are dying. Hate to sound like those old "kids will loose the ability to do math in their heads", anti-calculator types from my early days, but this is getting ridiculous.
highplainsdem
(56,258 posts)not loose
paleotn
(20,401 posts)So you're the grammar police? Sue me.
Niagara
(10,675 posts)I think you did a fantastic job, paleotn!
Humur seems ta bee lost on yalls aroun hear. OK I 
guess it's teh times. Everyboddie is so stresssed. 
Niagara
(10,675 posts)What a Capricorns finds humorous, others do not. I have to watch myself because I unintentionally piss people off.
However, it's seriously impolite for adults to correct other adults. I have seen a few name and grammar mistakes and I've sent the person a private message to let them know because that's how adults treat each other. Adults don't embarrass other adults publicly.
We have some seriously fucked up people here that will correct others grammar and then the person that does all the correcting makes continuous grammar errors. They come off as very Trumpian.
If you happen to spot a grammar error, kindly let that person know in private. Thank you.
blubunyip
(176 posts)OK so I forgot the sarcasm smilie-- 
my bad
Sorry to trigger your knee-jerk reaction. Next time I'll look out for thems who likes to pile on
paleotn
(20,401 posts)sometimes you can tell since there's no facial cues, body language and usually we don't know each other from Adam's off ox.
Spelling and grammar are not my thing. Not even close to my thing! My better half chuckles at me too sometimes. "How the hell did you think THAT was spelled correctly?!"
LuckyCharms
(20,055 posts)Publicly correcting spelling or grammar is frowned upon here.
As it should be.
Why bother correcting someone if it is apparent what they meant?
I know a very rude public figure who also falls back on "it was a joke ", or "that was sarcasm".
blubunyip
(176 posts)AI is very good at spelling (yay!), but wrong about facts sometimes (boo), and can be abused. Irony is implied.
It's a joke that pokes fun at the grammar police lightly. You can be sure that I think good spelling is not a greater virtue than good thinking. Bad spelling would suggest that an actual human wrote something, which is preferable to me. Maybe AI will incorporate some typos and bad spelling whenever it wants to appear really authentic.
Wow people are really defensive about bad spelling jokes! I did use the laughing guy but next time I'll remember 
sorry to offend
fujiyamasan
(208 posts)I use generative AI for limited use cases. It serves its purpose ok, but you have to know its limits and always verify.
But I kinda assume that if I’m posting something in a discussion forum, it’s with the intent and purpose of interacting with another human being.
Figarosmom
(5,906 posts)Escape
(210 posts)Now I am afraid to post anything clever here for fear of being accused of AI abuse.
(ps. I wouldn't know AI from E-I-E-I-O...)
David__77
(24,160 posts)
aggiesal
(10,054 posts)Other than that, I will not use A1 for any other purpose.
Bernardo de La Paz
(56,051 posts)"Education" Secretary Linda McMahon made the same mistake and was lampooned all over the internet about a month ago.
aggiesal
(10,054 posts)Now I'm think you are obviously being sarcastic as well.
Bernardo de La Paz
(56,051 posts)
forgotmylogin
(7,855 posts)We required that AI excerpts need to cite the source like a blog/journal/newspaper article and we will ban accounts that are trying to fool people by "masquerading" as a human - since an AI cannot legally agree to the Terms of Service/code of conduct. If an AI offends someone, how do we moderate it? We can run text through an AI bot checker to tell how likely it was to have been AI generated, and you do get an ear for it - most AI structures every answer like a book report with an opening thesis, supporting statements, then a conclusion.
I think AI has its use. It's kind of like using Wikipedia where it's a good place to start from to get an overview about a subject, but then you track down sources and read more about it.
Emile
(34,858 posts)they'll put in our brain to control our thoughts.
flashman13
(1,216 posts)I case you haven't noticed, the actual real car manufacturers have thrown in the towel on self driving cars.
Sympthsical
(10,634 posts)People were slow to adapt to it, afraid to try it, but now they're generally well thought of and I have friends who prefer it to dealing with a human drivers. It's interesting to observe.
Personally, I'm too skittish about that sort of thing. Particularly on a highway.
I do think the hype over-stepped the reality. The technology is still being worked on, and there was this idea that everything would be self-driving in 5-10 years. Which was completely unrealistic, and companies threw away a lot of money in all that hype.
I can see things like Waymo gaining more common use generationally. It just feels like one of those things that older people are extremely skeptical about, but those who grow up around the technology will become much less hesitant to use it.
And as far as, "We will replace truck drivers!" Yeah, maybe in 20 or 30 years. That shit is eight million lawsuits about to happen.
highplainsdem
(56,258 posts)the creatives whose work was stolen to train the AI should EVER use, unless forced to do so by their job or school.
If you're fine with the theft of the world's knowledge and intellectual property because genAI makes some things more convenient for you, it says a lot about you, but none of it is good.
It dumbs down and deskills users. It discourages critical thinking and creativity.
It's been a wrecking ball taken to education.
It's good for fraud, whether students cheating or people of any age pretending to have knowledge, skills, talent and even basic ideas they don't have.
Good for criminal fraud, which it's used for a lot.
Good for deepfakes, including porn - a popular use of it.
Good for data gathering and surveillance, which it's in demand for, from businesses and governments alike. Good for increasing the power and control exerted by those businesses and governments.
It's bad for the environment, with some AI companies having given up on previous environmental goals while they hope that AI will somehow save us from destroying the planet.
It's bad for our information ecosystem, polluting it with AI slop.
It's being peddled to businesses as a way to replace workers, not help them or shorten work weeks, and you're hearing less and less from the AI bros about the need for a UBI.
And it absolutely does not belong on a message board, where the purpose is communication between humans, not bots.
SheltieLover
(68,591 posts)Well said.
Ai should b we illegal imo.
MorbidButterflyTat
(3,029 posts)
TheProle
(3,375 posts)2. May or may not give an answer that's even relevant to the topic being discussed
3. May or may not be actively misinforming
How is this different from a substantial number of memes and tweets/social media posts that get reposted here?
Sympthsical
(10,634 posts)I think it just bugs me more when it's person-to-person rather than general content, as if faking a conversation.
It'd be like going to lunch with a friend to catch up, and whenever you asked a question or made a comment, they held up their phone in reply. At that point, it's like why did you bother hanging out at all?
Dumb is just baked into the Internet. But low effort faking feels a little insulting.
TheProle
(3,375 posts)LAS14
(15,197 posts)Or is it just gathering facts, or formatting a position?
Alice Kramden
(2,626 posts)I'm not sure whether it's in the rules - if not it should be
sl8
(16,517 posts)But, the AI posts I've noticed here were made with no indication that the words aren't those of the poster. That seems dishonest to me.
On the other hand, most of the social media posts I see here are pretty obviously not the work of the DU poster, even when they don't give explicit attribution to the original creator.
I do wish folks would be more careful to give accurate attribution for the material they post here, when possible, regardless of the source. I understand that sometimes it can be difficult to find the source and that's fine, but often it's trivially easy to find. Sometimes it seems like the poster just didn't bother to include the existing attribution in their cut & paste. I think I'm in the minority on this issue, though.
Celerity
(50,067 posts)The level of discernment on DU is far too often woeful (whether it is wilful or not I leave up to others to decide for themselves) at times.
FakeNoose
(37,486 posts)That's what has me puzzled.
If the words aren't your own, you must have gotten them somewhere. DU posters are required to include the link to where they got it, and AI such as Google or Chat aren't recognized sources.
sl8
(16,517 posts)Where did you see the DU requirement to provide links for material not your own? That would be pretty tough if you're quoting something from hardcopy media.
In any case, it's easy enough to give attribution to an AI generated piece. One easy way would be to use the DU "Excerpt" feature or even just quotation marks and preface the text with "Generated by Microsoft Copilot" or "Generated by Google Gemini" or something along those lines.
FakeNoose
(37,486 posts) 
SouthBayDem
(32,668 posts)Posting all sorts of unsolicited rated-R (NC-17?) material or malware-infected URLs.
Back when I was a kid experiencing stupid odd popups on my computer, I discovered a great informational website about what I was seeing: "Spyware, adware, malware, trojans, hijackers, and other Internet nuisances". That when GW Bush was in office! Luckily I still remember the URL from back then as the website was taken offline about 20 years ago.
LAS14
(15,197 posts)JCMach1
(28,698 posts)LAS14
(15,197 posts)hates it, loves it, disagrees with it. But has an opinion. AI cannot have an opinion. It can spout things that sound like opinions, but it can't entertain an opinion.
mucholderthandirt
(1,515 posts)Bye, DU, it was mostly good to know you. If this is allowed to continue, I'll be leaving. I hate "AI", it's all lies and stolen stuff, almost always wrong in every way. Anyone who can't write their own stuff, do their own research, has no right to post here.
Delphinus
(12,260 posts)And the timing of you bringing this up is great. Just this morning I read an article about AI and it said
It is an older story that came through a digest I receive; grateful it did because I missed it the first time around. Here's the link: https://medium.com/ai-ai-oh/the-one-and-only-reliable-way-to-spot-ai-today-ee4635bce44
The comments from teachers in other countries are enlightening as well.
IcyPeas
(23,621 posts)(The author is a sci-fi and fantasy author)
ecstatic
(34,784 posts)that much effort to respond to a post. By the time you are done prompting AI you could have already completed your post the old way. Not to mention, that would take all the fun out of message boards/forums.
But now that I think about it, I guess I could envision a scenario in which someone asks AI to clean up his/her post to make it grammatically correct, etc because we all know there are grammar / typo nazis here etc.
How about this: DU has been around for over 20 years. You could compare a user's old posts to posts made after mid-2024?
EarlG
(22,944 posts)You have no way of knowing whether someone's post was written by AI or not. You say you do, but you don't. Nobody does. That's why they're having all these problems with ChatGPT in academia right now.
You know that this is true, because you've been careful not to single anybody out. If you singled someone out, you'd be committing a personal attack. That's because you have absolutely no way to prove that someone's post was written by AI, so you'd be making a completely unfounded accusation. Do we really want a DU where everyone can go around accusing each other of using AI to write posts? I don't. That would get old very fast.
So instead you've got this sort of vague finger-pointing OP where you're not singling anyone out for using AI, but you're just saying that "people" are doing it. I don't really understand how this is helpful to the community, since it just seems to be making a lot of DUers paranoid.
I am not a fan of generative AI at all. I think for the most part it's a bunch of bullshit that is only going to make people who rely on it dumber and dumber over time. But the die is cast -- the technology is here, and I can't write a rule to stop it because there is literally no way to prove that anybody is using AI to write their posts.
And yet, before the advent of AI, anybody on DU could ask a partner or friend to edit their posts for clarity and spelling. They could even ask a partner or friend to write an entire post for them. Nobody would have been any the wiser -- but knowing that it was possible didn't seem to bother anybody.
I believe that most people who participate in this community do it not just because they want to read, but because they have something they want to say. I don't use AI to write my posts because I've been writing posts on here for 25 years and I enjoy doing it. I like using my brain. I think that the vast, vast majority of DUers feel the same way.
So for most people, it's a choice. Do you choose to believe that other DUers you're interacting with here are writing their own posts because they have something to say and they want to participate in the community, or do you choose to believe that other DUers aren't really people, and that you should be on high alert at all times because you might be talking to an AI?
I choose to believe the former. I think people participate here because they like using their brains and enjoy the process of engaging in written communication. But the purpose of this OP seems to be to encourage people to believe the latter, and the result is that you've created a long thread full of anxiety and paranoia. Since I'll also choose to believe that it was not your intention to do that, I don't really get the purpose of the OP.